Intermediate filaments attenuate stimulation‐dependent mobility of endosomes/lysosomes in astrocytes

Intermediate filament (IF) proteins upregulation is a hallmark of astrocyte activation and reactive gliosis, but its pathophysiological implications remain incompletely understood. A recently reported association between IFs and directional mobility of peptidergic vesicles allows us to hypothesize that IFs affect vesicle dynamics and exocytosis‐mediated astrocyte communication with neighboring cells. Here, we ask whether the trafficking of recycling vesicles (i.e., those fused to and then retrieved from the plasma membrane) and endosomes/lysosomes depends on IFs. Recycling vesicles were labeled by antibodies against vesicle glutamate transporter 1 (VGLUT1) and atrial natriuretic peptide (ANP), respectively, and by lysotracker, which labels endosomes/lysosomes. Quantitative fluorescence microscopy was used to monitor the mobility of labeled vesicles in astrocytes, derived from either wild‐type (WT) mice or mice deficient in glial fibrillary acidic protein and vimentin (GFAP^−/−Vim^−/−), the latter lacking astrocyte IFs. Stimulation with ionomycin or ATP enhanced the mobility of VGLUT1‐positive vesicles and reduced the mobility of ANP‐positive vesicles in WT astrocytes. In GFAP^−/−Vim^−/− astrocytes, both vesicle types responded to stimulation, but the relative increase in mobility of VGLUT1‐positive vesicles was more prominent compared with nonstimulated cells, whereas the stimulation‐dependent attenuation of ANP‐positive vesicles mobility was reduced compared with nonstimulated cells. The mobility of endosomes/lysosomes decreased following stimulation in WT astrocytes. However, in GFAP^−/−Vim^−/− astrocytes, a small increase in the mobility of endosomes/lysosomes was observed. These findings show that astrocyte IFs differentially affect the stimulation‐dependent mobility of vesicles. We propose that upregulation of IFs in pathologic states may alter the function of astrocytes by deregulating vesicle trafficking. © 2010 Wiley‐Liss, Inc.     

Plasminogen activator inhibitor-1 impairs plasminogen activation-mediated vascular smooth muscle cell apoptosis

The role of plasminogen activator inhibitor-1 (PAI-1) in vascular smooth muscle cell (VSMC) apoptosis mediated by plasminogen activation was studied with the use of aorticVSMC derived from mice with deficiency of PAI-1 (PAI-1 (-/-) ), tissue-type (t-PA (-/-) ) or urokinase-type (u-PA (-/-) ) plasminogen activator or from wildtype (WT) mice with corresponding genetic background. Plasminogen incubated with confluent VSMC was activated in a concentration-dependent and saturable manner for all four cell types, with maximal activation rates that were comparable for WT, u-PA (-/-) and t-PA (-/-) cells, but about two-fold higher for PAI-1 (-/-) cells. Plasminogen activation was impaired by addition of the lysine analogue 6-aminohexanoic acid, and by addition of t-PA and u-PA neutralizing antibodies, suggesting that it depends on binding to cell surface COOH-terminal lysine residues, and on plasminogen activator activity. Morphological alterations consistent with apoptosis were observed much earlier in PAI-1 (-/-) than in WT VSMC. Without addition of plasminogen, the apoptotic index was similar for all four cell types, whereas after incubation with physiological plasminogen concentrations, it was greater in PAI-1 (-/-) VSMC, as compared to WT, t-PA (-/-) or u-PA (-/-) VSMC. Furthermore, the apoptotic rate paralleled the release of plasmin. Thus, plasmin-mediated apoptosis of VSMC occurs via plasminogen activation by either t-PA or u-PA and is impaired by PAI-1.     

BQ788, an endothelin ET(B) receptor antagonist, attenuates stab wound injury-induced reactive astrocytes in rat brain.

Endothelins (ETs) are suggested to be involved in pathological or pathophysiological responses on brain injuries. In the present study, an involvement of ETs on activation of astrocytes in vivo was examined by using selective endothelin receptor antagonists. A stab wound injury on rat cerebral cortex increased immunoreactive ET-1 at the injured site. GFAP-positive [GFAP(+)] and vimentin-positive [Vim(+)] cells appeared at the injured site in 1 day to 2 weeks after the injury. A continuous infusion of BQ788, a selective ETB receptor antagonist, into cerebral ventricle (23 nmole/day) attenuated increase in the numbers of GFAP(+) and Vim(+) cells after the injury. FR139317, a selective ETA antagonist (23 nmole/day), slightly decreased the number of Vim(+) cells but not that of GFAP(+) cells. Increase in the number of microglia/macrophages by a stab wound injury, which was determined by Griffonia simplicifolia isolectin B4 staining, was not affected by BQ788 and FR139317. These results suggest that activation of glial ETB receptors is one of the signal cascades leading to reactive astrocytes on brain injuries.     

Aquaporin 4‐Dependent expression of glial fibrillary acidic protein and tenascin‐C in activated astrocytes in stab wound mouse brain and in primary culture

We previously reported that aquaporin 4 (AQP4) has a neuroimmunological function via astrocytes and microglial cells involving osteopontin. AQP4 is a water channel localized in the endofoot of astrocytes in the brain, and its expression is upregulated after a stab wound to the mouse brain or the injection of methylmercury in common marmosets. In this study, the correlation between the expression of AQP4 and the expression of glial fibrillary acidic protein (GFAP) or tenascin‐C (TN‐C) in reactive astrocytes was examined in primary cultures and brain tissues of AQP4‐deficient mice (AQP4/KO). In the absence of a stab wound to the brain or of any stimulation of the cells, the expressions of both GFAP and TN‐C were lower in astrocytes from AQP4/KO mice than in those from wild‐type (WT) mice. High levels of GFAP and TN‐C expression were observed in activated astrocytes after a stab wound to the brain in WT mice; however, the expressions of GFAP and TN‐C were insignificant in AQP4/KO mice. Furthermore, lipopolysaccharide (LPS) stimulation activated primary culture of astrocytes and upregulated GFAP and TN‐C expression in cells from WT mice, whereas the expressions of GFAP and TN‐C were slightly upregulated in cells from AQP4/KO mice. Moreover, the stimulation of primary culture of astrocytes with LPS also upregulated inflammatory cytokines in cells from WT mice, whereas modest increases were observed in cells from AQP4/KO mice. These results suggest that AQP4 expression accelerates GFAP and TN‐C expression in activated astrocytes induced by a stab wound in the mouse brain and LPS‐stimulated primary culture of astrocytes. © 2014 Wiley Periodicals, Inc.     

Motheaten (me/me) mice deficient in SHP-1 are less susceptible to focal cerebral ischemia

We have demonstrated previously that the protein tyrosine phosphatase SHP-1 seems to play a role in glial development and is upregulated in non-dividing astrocytes after injury. The present study examines the effect of loss of SHP-1 on the CNS response to permanent focal ischemia. SHP-1 deficient (me/me) mice and wild-type littermates received a permanent middle cerebral artery occlusion (MCAO). At 1, 3, and 7 days after MCAO, infarct volume, neuronal survival and cell death, gliosis, and inflammatory cytokine levels were quantified. SHP-1 deficient me/me mice display smaller infarct volumes at 7 days post-MCAO, increased neuronal survival within the ischemic penumbra, and decreased numbers of cleaved caspase 3+ cells within the ischemic core compared with wild-type mice. In addition, me/me mice exhibit increases in GFAP+ reactive astrocytes, F4-80+ microglia, and a concomitant increase in the level of interleukin 12 (IL-12) over baseline compared with wild-type. Taken together, these results demonstrate that loss of SHP-1 results in greater healing of the infarct due to less apoptosis and more neuronal survival in the ischemic core and suggests that pharmacologic inactivation of SHP-1 may have potential therapeutic value in limiting CNS degeneration after ischemic stroke.     

DJ‐1 immunoreactivity in human brain astrocytes is dependent on infarct presence and infarct age

DJ‐1 is a protein with anti‐oxidative stress and anti‐apoptotic properties that is abundantly expressed in reactive CNS astrocytes in chronic neurodegenerative disorders such as Parkinson's disease (PD), Alzheimer's disease (AD), and Pick's disease. Genetic mutations which eliminate DJ‐1 expression in humans are sufficient to produce an early‐onset form of familial PD, PARK7, suggesting that DJ‐1 is a critical component of the neuroprotective arsenal of the brain. Previous studies in parkinsonism/dementia brain tissues have revealed that reactive astrocytes within and surrounding incidentally identified infarcts were often robustly immunoreactive for DJ‐1, especially if the infarcts showed histological features consistent with older age. Given this, we sought to evaluate astrocytic DJ‐1 expression in human stroke more extensively, and with a particular emphasis on determining whether immunohistochemical DJ‐1 expression in astrocytes correlates with histological infarct age. The studies presented here show that DJ‐1 is abundantly expressed in reactive infarct region astrocytes in both gray and white matter, that subacute and chronic infarct region astrocytes are much more robustly DJ‐1+ than are acute infarct and non‐infarct region astrocytes, and that DJ‐1 staining intensity in astrocytes generally correlates with that of the reactive astrocyte marker GFAP. Confocal imaging of DJ‐1 and GFAP dual‐labelled human brain sections were used to confirm the localization to and expression of DJ‐1 in astrocytes. Neuronal DJ‐1 staining was minimal under all infarct and non‐infarct conditions. Our data support the conclusion that the major cellular DJ‐1 response to stroke in the human brain is astrocytic, and that there is a temporal correlation between DJ‐1 expression in these cells and advanced infarct age.     

Endothelin B receptors are expressed by astrocytes and regulate astrocyte hypertrophy in the normal and injured CNS

The ability of mammalian central nervous system (CNS) neurons to survive and/or regenerate following injury is influenced by surrounding glial cells. To identify the factors that control glial cell function following CNS injury, we have focused on the endothelin B receptor (ET(B)R), which we show is expressed by the majority of astrocytes that are immunoreactive for glial acid fibrillary protein (GFAP) in both the normal and crushed rabbit optic nerve. Optic nerve crush induces a marked increase in ET(B)R and GFAP immunoreactivity (IR) without inducing a significant increase in the number of GFAP-IR astrocytes, suggesting that the crush-induced astrogliosis is due primarily to astrocyte hypertrophy. To define the role that endothelins play in driving this astrogliosis, artificial cerebrospinal fluid (CSF), ET-1 (an ET(A)R and ET(B)R agonist), or Bosentan (a mixed ET(A)R and ET(B)R antagonist) were infused via osmotic minipumps into noninjured and crushed optic nerves for 14 days. Infusion of ET-1 induced a hypertrophy of ET(B)R/GFAP-IR astrocytes in the normal optic nerve, with no additional hypertrophy in the crushed nerve, whereas infusion of Bosentan induced a significant decrease in the hypertrophy of ET(B)R/GFAP-IR astrocytes in the crushed but not in the normal optic nerve. These data suggest that pharmacological blockade of astrocyte ET(B)R receptors following CNS injury modulates glial scar formation and may provide a more permissive substrate for neuronal survival and regeneration.     

Osteopontin is indispensable for activation of astrocytes in injured mouse brain and primary culture

ABSTRACT

Objectives: Osteopontin (OPN) is an inflammatory cytokine inducer involved in cell proliferation and migration in inflammatory diseases or tumors. To investigate the function of OPN in astrocyte activation during brain injury, we compared OPN-deficient (OPN/KO) with wild-type (WT) mouse brains after stab wound injury and primary culture of astrocytes.

Methods: Primary cultures of astrocytes were prepared from either WT or OPN/KO postnatal mouse brains. Activation efficiency of astrocytes in primary culture was accessed using Western blotting by examining the protein levels of glial fibrillary acidic protein (GFAP) and tenascin-C (TN-C), which are markers for reactive astrocytes, following lipopolysaccharide (LPS) stimulation. Furthermore, the stab wound injury on the cerebral cortex as a brain traumatic injury model was used, and activation of astrocytes and microglial cells was investigated using immunofluorescent analysis on fixed brain sections.

Results: Primary cultures of astrocytes prepared from WT or OPN/KO postnatal mouse brains showed that only 25% of normal shaped astrocytes in a flask were produced in OPN/KO mice. The expression levels of both GFAP and TN-C were downregulated in the primary culture of astrocytes from OPN/KO mice compared with that from WT mice. By the immunofluorescent analysis on the injured brain sections, glial activation was attenuated in OPN/KO mice compared with WT mice.

Discussion: Our data suggest that OPN is essential for proper astrocytic generation in vitro culture prepared from mouse cerebral cortex. OPN is indispensable for astrocyte activation in the mouse brain injury model and in LPS stimulated primary culture.

Abbreviations: AQP4: aquaporin 4; BBB: blood brain barrier; BrdU: bromo-deoxy uridine; CNS: central nervous system; GFAP: glial fibllirary acidic protein; IgG: immunoglobulin G; LPS: lipopolysaccharide; OPN: osteopontin; OPN/KO: osteopontin-deficient; TN-C: tenascin-C     

Effect of plasminogen activator inhibitor-1 deficiency on nutritionally-induced obesity in mice

Plasminogen activator inhibitor-1 (PAI-1) is the main physiological inhibitor of tissue-type (t-PA) and urokinase-type (u-PA) plasminogen activator. Recent studies in murine models have yielded apparently conflicting data on a potential role of PAI-1 in adipose tissue development and obesity. To reinvestigate this issue, we have rederived PAI-1 deficient (PAI-1(-/-)) and wild-type (WT) mice and generated true littermates in a 81.25% C57Bl/6: 18.75% 129 SV genetic background. Male 5-week-old PAI-1(-/-) and WT mice were kept on a high fat diet (20.1 kJ/g) for 15 weeks. Body weight gain was comparable for both genotypes, and at the time of sacrifice total body weights (39+/-1.1 versus 41 +/- 1.2 g) as well as the weights of subcutaneous (SC, 1,520 +/- 110 versus 1,480 +/- 110 mg) adipose tissue were not significantly different. In contrast, the gonadal (GON, 1,900 +/- 43 versus 1,510 +/- 86 mg, p < 0.005) tissue mass was larger in PAI-1(-/-) mice. Plasma levels of insulin, leptin, glucose, triglycerides, total, HDL and LDL cholesterol were comparable for both genotypes. Immunohistochemical analysis of SC and GON adipose tissues did not reveal differences in adipocyte size or number between both genotypes, whereas blood vessel density was also comparable for GON fat but lower in SC fat of WT mice. Thus, this study in littermate mice on high fat diet did not reveal an effect of PAI-1 deficiency on body weight, and a differential effect on SC and GON adipose tissue.     

Focal cerebral ischemia upregulates SHP-1 in reactive astrocytes in juvenile mice

The role of the tyrosine phosphatase SHP-1 in the hematopoietic system has been well studied; however, its role in the central nervous system (CNS) response to injury is not well understood. Previous studies in our laboratory have demonstrated increased immunoreactivity for SHP-1 in a subset of reactive astrocytes that do not appear to enter the cell cycle following deafferentation of the chicken auditory brainstem. In order to determine whether mammalian astrocytes also upregulate SHP-1 immunoreactivity following CNS injury, a mouse model of focal cerebral ischemia was utilized to study SHP-1 expression. The brains of 3-week-old mice were analyzed at four time points following permanent middle cerebral artery occlusion (MCAO): 1, 3, 7, and 14 days. Our results demonstrate consistent infarct volumes within surgical groups, and infarct volumes decrease as a function of time from 1 day (maximum infarct volume) to 14 days (minimum infarct volume) post-MCAO. In addition, SHP-1 protein levels are upregulated following cerebral ischemia and this increase peaks at 7 days post-MCAO. Analysis of confocal images further reveals that immunoreactivity for SHP-1 occurs predominantly in GFAP+ reactive astrocytes, although a small percentage of F4-80+ microglia are also double labeled for SHP-1 at early times post-MCAO. These SHP-1+ reactive astrocytes do not appear to enter the cell cycle (as defined by PCNA immunoreactivity), confirming our previous studies in the avian auditory brainstem. These results suggest that SHP-1 plays an important role in the regulation of glial activation and proliferation in the ischemic CNS.     

Growth inhibition of vascular smooth muscle cells derived from urokinase receptor (u-PAR)-deficient mice in the presence of carcinoma cells

The growth rate of vascular smooth muscle cells (VSMCs), which were derived from aorta of mice deficient in the fibrinolytic factors tissue-type plasminogen activator (t-PA(-/-)), urokinase (u-PA(-/-)), u-PA receptor (u-PAR(-/-)) and type 1 plasminogen activator inhibitor (PAI-1(-/-)), as well as wild-type (WT) mice, was investigated in the presence of mouse melanoma cells (B16). In the VSMCs cultured with a basal medium supplemented with 10% fetal calf serum (FCS), there was no difference in the growth rate among the gene-lacking VSMCs and WT VSMCs, indicating that these fibrinolytic factors were not involved in the FCS-mediated cell proliferation. On the other hand, when these VSMCs were cultured with B16 cells in either the mixed culture or a double-chamber, only u-PAR(-/-) VSMCs showed a significantly lower growth rate. In addition, these suppressive effects on u-PAR(-/-) VSMCs were also observed in the presence of B16-derived conditioned medium (B16/CM). The growth rate of all the VSMCs except u-PAR(-/-) VSMCs was increased in the presence of B16/CM. The degree of the increase in cell number was comparable to that obtained with FCS. These effects on growth activity were partially associated with the levels of mitogen-activated protein kinase (MAPK, p42/p44) activity. The findings suggest that u-PAR plays an important role in the proliferative response of VSMCs and that without u-PAR, there is no intracellular signaling for cell proliferation.     

The intermediate filament GFAP is important for the control of experimental murine Staphylococcus aureus-induced brain abscess and Toxoplasma encephalitis

The functional role of astrocytes exerted via their intermediate protein glial fibrillary acidic protein (GFAP) in CNS infections was studied in Staphylococcus aureus-induced brain abscess. Compared to wild type (WT) mice, GFAP(0/0) mice developed larger and more poorly demarcated inflammatory lesions paralleled by a significantly increased intracerebral bacterial load, a diffuse leukocytic infiltration of the contralateral hemisphere, purulent ventriculitis, vasculitis, and severe brain edema. These observations were correlated with the lack of a bordering function of activated astrocytes that strongly upregulated their GFAP expression in the abscess surrounding of WT mice. Clinically important, this lack of restriction of inflammation markedly aggravated the course of disease with manifestation of seizures and a severe weight loss in GFAP(0/0) mice. These data were paralleled by observations in the model of Toxoplasma encephalitis (TE) during which the intracerebral parasitic load was significantly increased. Moreover, tachyzoite-induced tissue necrosis was exclusively found in the brains of GFAP(0/0) mice in chronic TE. Collectively, these findings delineate a host defense function of astrocytes via restricting pathogenic spread and multiplication within the CNS, thereby contributing to the protection of the highly vulnerable brain parenchyma.     

Urokinase‐type plasminogen activator modulates mammalian circadian clock phase regulation in tissue‐type plasminogen activator knockout mice

Glutamate phase shifts the circadian clock in the mammalian suprachiasmatic nucleus (SCN) by activating NMDA receptors. Tissue‐type plasminogen activator (tPA) gates phase shifts by activating plasmin to generate m(ature) BDNF, which binds TrkB receptors allowing clock phase shifts. Here, we investigate phase shifting in tPA knockout (tPA^?/?; B6.129S2‐Plat^tm1Mlg/J) mice, and identify urokinase‐type plasminogen activator (uPA) as an additional circadian clock regulator. Behavioral activity rhythms in tPA^?/? mice entrain to a light‐dark (LD) cycle and phase shift in response to nocturnal light pulses with no apparent loss in sensitivity. When the LD cycle is inverted, tPA^?/? mice take significantly longer to entrain than C57BL/6J wild‐type (WT) mice. SCN brain slices from tPA^?/? mice exhibit entrained neuronal activity rhythms and phase shift in response to nocturnal glutamate with no change in dose‐dependency. Pre‐treating slices with the tPA/uPA inhibitor, plasminogen activator inhibitor‐1 (PAI‐1), inhibits glutamate‐induced phase delays in tPA^?/? slices. Selective inhibition of uPA with UK122 prevents glutamate‐induced phase resetting in tPA^?/? but not WT SCN slices. tPA expression is higher at night than the day in WT SCN, while uPA expression remains constant in WT and tPA^?/? slices. Casein‐plasminogen zymography reveals that neither tPA nor uPA total proteolytic activity is under circadian control in WT or tPA^?/? SCN. Finally, tPA^?/? SCN tissue has lower mBDNF levels than WT tissue, while UK122 does not affect mBDNF levels in either strain. Together, these results suggest that either tPA or uPA can support photic/glutamatergic phase shifts of the SCN circadian clock, possibly acting through distinct mechanisms.     

Chronic relapsing experimental allergic encephalomyelitis (CREAE) in plasminogen activator inhibitor-1 knockout mice: the effect of fibrinolysis during neuroinflammation

During neuroinflammation in multiple sclerosis (MS) fibrinogen, not normally present in the brain or spinal cord, enters the central nervous system through a compromised blood–brain barrier. Fibrin deposited on axons is ineffectively removed by tissue plasminogen activator (tPA), a key contributory factor being the upregulation of plasminogen activator inhibitor-1 (PAI-1). Aims: This study investigated the role of PAI-1 during experimental neuroinflammatory disease. Methods: Chronic relapsing experimental allergic encephalomyelitis (CREAE), a model of MS, was induced with spinal cord homogenate in PAI-1 knockout (PAI-1^−/−) and wild type (WT) mice, backcrossed onto the Biozzi background. Results: Disease incidence and clinical severity were reduced in PAI-1^−/− mice, with animals developing clinical signs significantly later than WTs. Clinical relapses were absent in PAI-1^−/− mice and the subsequent reduction in neuroinflammation was coupled with a higher capacity for fibrinolysis in spinal cord samples from PAI-1^−/− mice, in association with increased tPA activity. Axonal damage was less apparent in PAI-1^−/− mice than in WTs, implicating fibrin in both inflammatory and degenerative events during CREAE. Conclusions: PAI-1 is a potential target for therapy in neuroinflammatory degenerative diseases, allowing effective fibrin removal and potentially reducing relapse rate and axonal damage.     

Glial HMOX1 expression promotes central and peripheral α-synuclein dysregulation and pathogenicity in parkinsonian mice

α-Synuclein is a key player in the pathogenesis of Parkinson disease (PD). Expression of human heme oxygenase-1 (HO-1) in astrocytes of GFAP.HMOX1 transgenic (TG) mice between 8.5 and 19 months of age results in a parkinsonian phenotype characterized by neural oxidative stress, nigrostriatal hypodopaminergia associated with locomotor incoordination, and overproduction of α-synuclein. We identified two microRNAs (miR-), miR-153 and miR-223, that negatively regulate α-synuclein in the basal ganglia of male and female GFAP.HMOX1 mice. Serum concentrations of both miRNAs progressively declined in the wild-type (WT) and GFAP.HMOX1 mice between 11 and 19 months of age. Moreover, at each time point surveyed, circulating levels of miR-153 were significantly lower in the TG animals compared to WT controls, while α-synuclein protein concentrations were elevated in erythrocytes of the GFAP.HMOX1 mice at 19 months of age relative to WT values. Primary WT neurons co-cultured with GFAP.HMOX1 astrocytes exhibited enhanced protein oxidation, mitophagy and apoptosis, aberrant expression of genes regulating the dopaminergic phenotype, and an imbalance in gene expression profiles governing mitochondrial fission and fusion. Many, but not all, of these neuronal abnormalities were abrogated by small interfering RNA (siRNA) knockdown of α-synuclein, implicating α-synuclein as a potent, albeit partial, mediator of HO-1's neurodystrophic effects in these parkinsonian mice. Overexpression of HO-1 in stressed astroglia has previously been documented in the substantia nigra of idiopathic PD and may promote α-synuclein production and toxicity by downmodulating miR-153 and/or miR-223 both within the CNS and in peripheral tissues.     

Glutamate metabolism is down-regulated in astrocytes during experimental allergic encephalomyelitis

Experimental allergic encephalomyelitis (EAE) was induced in SJL/J mice by adoptive transfer of MBP-reactive T cells in order to investigate the role of astrocytes in pathology. GFAP protein and mRNA expression (analyzed using semi-quantitative Western blot and RT-PCR techniques) were upregulated in the spinal cord of mice, which had developed a complete paralysis of hind- and fore-limbs and tail (grade 4 EAE), thus establishing that reactive gliosis occurred under these experimental conditions. Within the same samples and using similar techniques, we found that glutamine synthetase (GS) and glutamate dehydrogenase (GDH) expression were dramatically reduced. These two astrocytic enzymes are responsible for degradation of glutamate, the most abundant excitatory neurotransmitter in the brain. Since elevated levels of glutamate may be neurotoxic, we propose that the decreased capacity of astrocytes to metabolize glutamate may contribute to EAE pathology. GLIA 20:79–85, 1997. © 1997 Wiley-Liss, Inc.     

Astrocytes support hippocampal-dependent memory and long-term potentiation via interleukin-1 signaling

Recent studies indicate that astrocytes play an integral role in neural and synaptic functioning. To examine the implications of these findings for neurobehavioral plasticity we investigated the involvement of astrocytes in memory and long-term potentiation (LTP), using a mouse model of impaired learning and synaptic plasticity caused by genetic deletion of the interleukin-1 receptor type I (IL-1RI). Neural precursor cells (NPCs), derived from either wild type (WT) or IL-1 receptor knockout (IL-1rKO) neonatal mice, were labeled with bromodeoxyuridine (BrdU) and transplanted into the hippocampus of either IL-1rKO or WT adult host mice. Transplanted NPCs survived and differentiated into astrocytes (expressing GFAP and S100β), but not to neurons or oligodendrocytes. The NPCs-derived astrocytes from WT but not IL-1rKO mice displayed co-localization of GFAP with the IL-1RI. Four to twelve weeks post-transplantation, memory functioning was examined in the fear-conditioning and the water maze paradigms and LTP of perforant path-dentate gyrus synapses was assessed in anesthetized mice. As expected, IL-1rKO mice transplanted with IL-1rKO cells or sham operated displayed severe memory disturbances in both paradigms as well as a marked impairment in LTP. In contrast, IL-1rKO mice transplanted with WT NPCs displayed a complete rescue of the impaired memory functioning as well as partial restoration of LTP. These findings indicate that astrocytes play a critical role in memory functioning and LTP, and specifically implicate astrocytic IL-1 signaling in these processes. The results suggest novel conceptualization and therapeutic targets for neuropsychiatric disorders characterized by impaired astrocytic functioning concomitantly with disturbed memory and synaptic plasticity.     

Tissue-type plasminogen activator-plasmin-BDNF modulate glutamate-induced phase-shifts of the mouse suprachiasmatic circadian clock in vitro

The mammalian circadian clock in the suprachiasmatic nucleus (SCN) maintains environmental synchrony through light signals transmitted by glutamate released from retinal ganglion terminals. Brain-derived neurotrophic factor (BDNF) is required for light/glutamate to reset the clock. In the hippocampus, BDNF is activated by the extracellular protease, plasmin, which is produced from plasminogen by tissue-type plasminogen activator (tPA). We provide data showing expression of proteins from the plasminogen activation cascade in the SCN and their involvement in circadian clock phase-resetting. Early night glutamate application to SCN-containing brain slices resets the circadian clock. Plasminogen activator inhibitor-1 (PAI-1) blocked these shifts in slices from wild-type mice but not mice lacking its stabilizing protein, vitronectin (VN). Plasmin, but not plasminogen, prevented inhibition by PAI-1. Both plasmin and active BDNF reversed α₂-antiplasmin inhibition of glutamate-induced shifts. α₂-Antiplasmin decreased the conversion of inactive to active BDNF in the SCN. Finally, both tPA and BDNF allowed daytime glutamate-induced phase-resetting. Together, these data are the first to demonstrate expression of these proteases in the SCN, their involvement in modulating photic phase-shifts, and their activation of BDNF in the SCN, a potential 'gating' mechanism for photic phase-resetting. These data also demonstrate a functional interaction between PAI-1 and VN in adult brain. Given the usual association of these proteins with the extracellular matrix, these data suggest new lines of investigation into the locations and processes modulating mammalian circadian clock phase-resetting.